Remove data copying for internal component variables, derivatives, and residuals in PowerElectronics

[alexander-novo]

Description

Partially addresses #353.

Removes data copying between the system and components for internal variables, derivatives, and residuals when evaluating residuals and Jacobians.

Proposed changes

CircuitComponent now has three pointers:

• int_: a pointer to the component's internal variables in the system vector
• intp_: a pointer to the component's internal derivatives in the system vector
• int_f_: a pointer to the component's internal residuals in the system vector

PowerElectronicsModel::distributeVectors() and PowerElectronicsModel::evaluateResidual() have been modified to only copy variables between y_ and f_ if they are external. Therefore, components must use the three new pointers above to correctly calculate residuals and Jacobians.

Checklist

• All tests pass.
• Code compiles cleanly with flags -Wall -Wpedantic -Wconversion -Wextra.
• The new code follows GridKit™ style guidelines.
• There are unit tests for the new code.
• The new code is documented.
• The feature branch is rebased with respect to the target branch.
• I have updated CHANGELOG.md to reflect the changes in this PR. If this is a minor PR that is part of a larger fix already included in the file, state so.

Further comments

This solution won't work quite right for system objects which are components in another system. Maybe we want some sort of flag that checks for this situation and doesn't allocate a y_ or f_ vector?

[alexander-novo]

@pelesh I'm not really sure what to do about the enzyme example, since the DG model's residual has diverged from what's given in that example. Let me know what you think is best.

[pelesh]

@pelesh I'm not really sure what to do about the enzyme example, since the DG model's residual has diverged from what's given in that example. Let me know what you think is best.

Let's merge the other power electronics PRs, first.

My understanding is that only variables were reordered and equations remained the same. In that case, it should not be difficult to fix the answer key for the Enzyme test.

[pelesh]

I set this to draft for now until we merge other related PRs.

[alexander-novo]

@pelesh I'm not sure the DistributedGeneratorTest makes sense. The test expects to evaluate the residual of a reference DG where the delta variable is non-zero, which is now no longer possible and I don't think should be a supported feature of the model?

[alexander-novo]

Well I just changed it to be a test of a non-reference generator, and edited the answer key to match.

[pelesh]

This is very nice work for what I could see thus far. I would only avoid calling internal variables int_ becayuse in C-speak "int" is widely recognized as "integer" and int_ is too close for comfort to the keyword int.

How about y_int_?